McCormack Versus Udall

• DOI: 10.1177/1532673X05277135
• Date: 01 January 2006
• Published date: 01 January 2006
• Author: Matthew N. Green
• Subject Matter: Articles

American Politics Research
Volume 34 Number 1
10.1177/1532673X05277135
American Politics Research
Green / McCormack Versus Udall
January 2006 3-21
© 2006 Sage Publications
McCormack Versus Udall
10.1177/1532673X05277135
http://apr.sagepub.com
hosted at
Explaining Intraparty Challenges
http://online.sagepub.com
to the Speaker of the House
Matthew N. Green
Catholic University of America
Since the mid-1970s, little research has been undertaken on the subject of
intraparty leadership challenges in Congress. In this article, I introduce a the-
ory to explain who will most likely support a challenge to an incumbent
speaker of the House. I then test this theory for one such challenge in particular,
against Speaker John McCormack in 1969, using newly obtained archival data
related to the contest. The results suggest that legislators are most likely to sup-
port a challenge when it promises to satisfy one or more “multiple goals.” The
findings also reveal that membership in the same entering class in Congress
plays an important and previously unacknowledged causal role in determining
support for a challenge to the speaker.
Keywords: U.S. Congress; congressional leadership; speaker of the House;
Morris Udall; John McCormack; leadership elections; U.S. House of Repre-
sentatives; Democratic Caucus
Notsincethe1970s havecongressionalscholarsexaminedindepththe
subject of intraparty congressional leadership challenges: that is,
moments when an incumbent party leader in Congress is challenged for the
office by another legislator from within his or her own party. These chal-
lenges are an important feature of legislative politics for several reasons:
They have significant effects on the political careers of both incumbents and
challengers; they can alter the policy direction and political strategy of the
party in question; and they illustrate the separate interests, goals, and prefer-
ences that divide incumbent parties (Nelson, 1977, p. 928; Peabody, 1967,
p. 675).1 Furthermore, such challenges, at least with respect to the speaker of
Author’s Note: My thanks to Frances Lee, David Mayhew, and the anonymous reviewers for
their helpful comments and suggestions; and to Roger Myers (Special Collections, University of
Arizona Library) and Sean Noel and Charles Niles (Howard Gotlieb Archival Research Center,
Boston University) for their assistance in obtaining historical data.
3

---

4
American Politics Research
the House, have become more frequent over the past 30 years, in spite of the
many risks associated with them and their slim odds of success.2
Earlier studies of this topic suggest some explanations for why party
members support such challenges. In a 1967 article and a subsequent 1976
book, Robert Peabody identified important differences between majority
Democrats and minority Republicans, as well as between the House and the
Senate, in observed patterns of leader replacement. Peabody (1976) argued
that a number of conditions may matter but emphasized the role of internal
factors (particularly the characteristics of the incumbent leader) and the
influence of congressional election results. Following Peabody’s analysis,
Garrison Nelson (1977) took a broader historical perspective, looking at con-
gressional leadership change from 1789 through 1977, and also found im-
portant differences between the two main congressional parties, including
the nature of formal systems of succession and the frequency of successful
challenges.
In this analysis, I seek to build on these prior studies, focusing on intra-
party leadership change for one office in particular: the speaker of the House
of Representatives. Because the speaker is the top party leader in the House,
challenges to his or her office are of special significance for the policy direc-
tion and leadership structure of Congress. I also focus on a specific research
question: Who in the majority party is most likely to support a challenge to an
incumbent speaker when it ends in a formal vote? I argue that a “multiple
goals” approach to congressional politics, which recognizes that legisla-
tors have several distinct goals that they may wish to achieve, provides the
necessary analytical leverage to explain who supports an intraparty leader-
ship challenge. More specifically, I show that legislators’ characteristics—
including ideological preferences, age, and one’s electoral cohort—
associated with the goals of internal influence and desired policy enactment
dictate the likelihood of opposing an incumbent speaker.
The outline of the discussion is as follows. I first introduce a theory to
explain who would be most likely to cast a balloted vote in support of an
intraparty challenge to the speaker. I then use previously unused archival data
to test this theory and alternative hypotheses for one case in particular: the
1969 challenge against Speaker John McCormack (D-MA), one of the most
successful open contests against any incumbent speaker in the past four
decades.3 I conclude with comments on the applicability of my findings to
other intraparty leadership challenges in Congress.

---

Green / McCormack Versus Udall
5
Explaining Leadership Challenges
A few scholars have explicitly or implicitly suggested theories to explain
whether an individual legislator would support a leadership challenge within
his or her party. In perhaps the most comprehensive study of congressional
leadership change to date, Robert Peabody (1976) identified a variety of both
external and internal factors that can explain who is chosen as a party leader
and, by implication, who is most likely to vote for a challenger against an
incumbent party leader, including the speaker.4 Peabody argued that internal
factors, including the ideology, constituency, regional background, and
seniority of candidates, were generally more significant than outside factors,
such as presidential influence, the media, or interest group activity (pp. 469,
494-497).5 Two factors in particular, he argued, were consistently important
in explaining leadership change: congressional election results and the per-
sonalities and skills of the incumbent and challenger(s) (Peabody, 1967, pp.
687-689; Peabody, 1976, pp. 299-300, 490, 498).6 However, Peabody shied
away from posing a single overarching theory to explain leadership change,
arguing that different factors come into play in different cases.
A second approach to congressional leadership—and, by extension, lead-
ership selection—focuses on the idea that legislators, or the “principals,”
have certain expectations of party leaders, who serve as their “agents”
(Jones, 1968, p. 618; Rohde & Shepsle, 1987; Sinclair, 1995). In perhaps the
most explicit elucidation of this approach, Barbara Sinclair (1995) draws
from the observations of Richard Fenno (1973) to argue that party leaders are
agents chosen by their caucus to fulfill one or more of three goals: reelection
to Congress, the enactment of desired policy, and influence within the House.
The implication is that those caucus members whose expectations have been
least satisfied by the incumbent speaker are most likely to support the ouster
of that speaker in the event of an intraparty challenge (Fenno, 1973; Sinclair,
1995, p. 18). This multiple-goals approach to principal-agent theory rightly
emphasizes the electoral connection between a party leader and his or her
caucus, while narrowing the number of factors relevant to leadership change.
However, it does not explain what makes one goal more important than
another for a particular challenge or how they relate to the internal and
external factors identified by Peabody (1976).
I incorporate the insights of both Peabody (1976) and Sinclair (1995) into
a more precise theoretical framework. As Peabody observes, intraparty lead-
ership challenges can involve a variety of factors that vary in importance
across cases. However, given the importance of reelection, policy, and inter-
nal influence for legislators—and the expectation that party leaders are
responsible, at least indirectly, for helping legislators achieve those goals—

---

6
American Politics Research
the factors at play in a leadership challenge should be closely associated with
one or more of these three goals.
But not all legislator goals will matter equally in determining vote choice.
As with elections in general, intraparty leadership challenges will vary in the
quality and background of the candidates, the issues raised during the cam-
paign, and the efforts taken to win votes. It is this variation in context, I argue,
that determines the causal importance of a given goal. In particular, the prior
records of the candidates and the terms of the campaign will determine
whether a particular goal is salient. The prior records of the incumbent and
challenger include their reputations within the party, their past leadership
decisions, and scarce goods they have (or have not) distributed to party mem-
bers, whereas the terms of the campaign include how the candidates frame
the campaign or to whom they promise particular scarce benefits in exchange
for votes. In short, this goal-salience theory of intraparty leadership chal-
lenges holds that only certain legislator goals, made relevant by the context
of a given leadership challenge, will best predict the likelihood that a caucus
member will support or oppose that challenge.
How do the three goals of legislators translate into specific, measurable
interests or characteristics of individual majority party members? The first
goal, reelection, is usually considered important for leadership challenges
only...